Cyber Attacks Detection Based on Generative Adversarial Networks

The development of the Internet of things (IoT) and other modern technologies brings not only great convenience and intelligence to society but also the severe problem of cyber security. In recent years, a large number of network attacks, especially botnet and distributed denial of service (DDoS) attacks which are difficult to detect, have seriously affected the normal operation of social work. Solutions and methods based on artificial intelligence techniques are therefore proposed to help detect cyberattacks. In this paper, the application of GAN in the field of network security is discussed. GAN has been one of the most significant deep learning models in recent years. It consists of a generative model and a discriminative model which can be combined to produce a dynamic game system. GAN has been applied in various fields such as image generation, speech processing, data enhancement, and cyberattack detection. This paper demonstrates the basic working principle and infrastructure of GAN, and also focuses on how the GAN model assists with cyber intrusion detection.